重新审视基于视觉推理泛化性的长链式推理的必要性
Revisiting the Necessity of Lengthy Chain-of-Thought in Vision-centric Reasoning Generalization
AI总结：
本文研究了不同链式推理设计对视觉推理泛化能力的影响，发现简洁的链式推理在不同迷宫规模中表现最佳，提出'短即长'效应。
摘要：
我们研究了不同链式推理（CoT）设计如何影响视觉语言模型（VLMs）获得可推广的视觉推理能力。尽管链式推理数据，特别是长或视觉链式推理如“图像思考”已被广泛用于监督中间推理，但仍然不清楚为什么特定的链式推理设计有助于提升，哪些真正支持可推广的推理。为了系统评估这一点，我们专注于一个受控的迷宫解决基准，其中推理规则完全视觉化，难度可以通过网格大小调节，且所有中间步骤都可以自动生成。使用Qwen2.5-VL-7B在标准的SFT-then-RL流程下，我们比较了三种代表性的链式推理格式：语言链式推理、定位链式推理（带有空间坐标轨迹）和视觉链式推理（带有图像操作）。我们的实验发现，视觉和更长的链式推理主要加速收敛，但不会提升最终性能上限；简洁的链式推理仅包含必要的定位步骤表现优于更长的轨迹；并且，令人惊讶的是，仅保留最小定位的链式推理在不同迷宫大小上泛化效果最佳。我们进一步在其他视觉中心任务上验证了这些见解。这些发现突显了“短即长”的效应，并为构建更具可推广性的SFT数据集提供了实用指导。
英文摘要：
We study how different Chain-of-Thought (CoT) designs affect the acquisition of the generalizable visual reasoning ability in vision-language models (VLMs). While CoT data, especially long or visual CoT such as "think with image", has been widely used to supervise intermediate reasoning, it remains unclear why specific CoT designs help and which ones truly support generalizable reasoning. To systematically evaluate this, we focus on a controlled maze-solving benchmark where reasoning rules are fully visual, difficulty can be tuned by grid size, and all the intermediate steps can be automatically generated. Using Qwen2.5-VL-7B under a standard SFT-then-RL pipeline, we compare three representative CoT formats: Language CoT, Grounding CoT (with spatial coordinate trajectories), and Visual CoT (with image manipulations). Our experiments reveal that visual and longer CoT mainly accelerate convergence but do not lift the final performance ceiling; concise CoT containing only essential grounding steps outperforms longer traces; and, strikingly, CoT retaining only the minimal grounding results generalizes best across different maze sizes. We further validate these insights on other vision-centric tasks. These findings highlight a "short is long" effect and provide practical guidance for constructing more generalizable SFT datasets for visual reasoning.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
